TSMC Hacked, Vibes Down Outlook Down, 41 Days Till Christmas, Nvidia v2

0:00

You're watching TVN.

0:02

>> Today is Friday, November 13th, 2026.

0:02

We are live from the TVPN Ultradome, the temple of technology, the fortress of finance, >> the capital of capital.

0:12

>> I feel like your voice is coming back a little bit. >> Maybe.

0:14

>> Let me tell you about ramp. com. Time is money. Save both.

0:16

It used to use corporate cards, bill pay, accounting, and a whole lot more all in one place.

0:22

Uh, I think we're I think we're over the hump.

0:24

over the hump of both our sickness and >> very weird sickness because we were sick like a long time ago. Yeah.

0:32

>> And then >> like you know long road long road back.

0:37

>> Uh >> well speaking of roads, let's check in on the road to Christmas. >> Perfect.

0:40

>> How many days until Christmas?

0:42

>> We need to count it down.

0:42

That's what you come to TBTN for.

0:44

We have you covered on the road to Christmas.

0:46

41 [music] days until Christmas.

0:48

Sam Alman 4 days 41 years old. Not his birthday today. >> Not today.

0:55

>> Just generally he's 41. >> Random. >> There you go. >> Number association. >> Fun fun fact.

1:00

Uh anyway, we are uh through most of the Nvidia news.

1:06

It feels like uh the the week is winding down.

1:09

It really crescendoed in the middle of the week.

1:11

Uh yesterday was uh was was crazy.

1:14

the most fallout I think of everything as Jensen resigned as CEO and uh doesn't even hold a board position anymore.

1:22

Uh so there's a few lingering um stories that we should catch people up on, but we do want to move on from this story because there's so much else to talk about. Um >> is there? >> Yes, there is.

1:32

There is the mansion section.

1:34

[laughter] >> Um yeah, know it was funny.

1:37

Uh, I saw a comment from from somebody I think this morning saying, "All right, guys.

1:41

You've been talking about the Jensen story for nine days. We get it." And uh, >> nine days.

1:48

I don't think it's been >> I I know I know you want to talk us to talk about other stuff, but uh, I think it's ultimately the biggest story >> y >> in technology in my lifetime.

2:01

In many ways, the other thing is it has the most potential downstream impacts on like the entire >> all the private everything happening in the private markets um public markets.

2:13

Uh we're >> we're not done.

2:14

We're going to keep talking about it.

2:16

>> The other thing is that uh when one of these big all-encompassing timeline takeover current thing stories happens like people do pull back on creating more news.

2:25

So uh on Monday we were sort of like whoa weird timing uh for Anthropic to drop Fable 5. 2.

2:30

two, it's going to be hard for them to break through with that model.

2:34

Whereas, if it's a really slow news week, a bunch of people are going to be sharing, here's what I built with it, here are the benchmarks, let's debate uh AGI timelines, safety, blah blah blah.

2:42

It was kind of just like, here's the release card, go for it.

2:46

And uh it didn't break through nearly as much.

2:49

Uh and that's something that they clearly had planned in preparation for the IPO next week.

2:54

they weren't going to cancel that.

2:56

But many other companies are pushing back releases right now because of the Jensen story because if you have a funding announcement or some product launch, you know that it's going to be harder to break through.

3:06

So that also gives us a little bit less to talk about.

3:09

But fortunately, the good folks at the Wall Street Journal, they're still publishing the mansion section and so we will have plenty to talk about.

3:16

But um first we should do a little review of last night.

3:19

Jordy, how did you enjoy GQ man of the year? Really fun. >> Great party. >> Really fun.

3:28

Uh >> first red carpet ever. >> Yes. >> Yes.

3:32

>> We should get a red carpet in here.

3:32

I feel like for the for the guests to enter.

3:35

>> Don't you think that's underrated?

3:35

We have a couple black carpets, but >> um but it was really fun.

3:38

Honestly, we spent the entire time going around hanging out with all the media people. >> Yeah.

3:45

Yeah, the business media people.

3:47

>> It's not really our scene, but we had a really, >> really great time.

3:49

Uh Louis Vuitton was kind enough to dress us. >> Yes, thank you. Louis Vuitton. Fantastic.

3:56

>> We felt uh we felt incredible out there.

3:58

>> Yeah, it was remarkable. Uh it was a movie.

4:01

>> Yeah, we needed Chad's saying we need a green carpet. >> Green carpet. Okay.

4:03

Yeah, maybe we do a green carpet.

4:06

Yeah, Dylan was there with us. We had a lot of fun. Uh anyway, not much.

4:11

Uh oh, the one the one interesting takeaway was hearing from folks in Hollywood about their take on the Jensen story.

4:18

So the Jensen story has broken through.

4:22

I mean we've discussed it as breaking through in the sense that I will see uh Instagram creators and reals creators who normally their their domain is just uh impressions or day in the life or doing something non- techreated.

4:37

They're all taking the Jensen story and doing their version of content around it.

4:42

Uh but it's also broken through to Hollywood media broadly.

4:45

It's not a pure tech story by any means. We we know this.

4:50

But uh what was interesting was that uh we were talking to one individual who was saying that um this signaled the end of AI in some way and that it was this ironic beautiful story that something so human affair would bring down the techno machine, the relentless march of techno capitalism. >> Yeah. That Yeah.

5:12

His takeaway was um but that it was completely over. Yes. like AI is done for. >> Yeah.

5:19

>> Um which was uh which was fascinating.

5:22

>> We had to be like we got some bad news for you.

5:24

This potentially concentrates power in just a smaller handful of companies and might [snorts] might actually create create more acceleration in in a weird twist.

5:32

Um and certainly uh nothing is is slowing down in terms of model progress or or AI diffusion.

5:38

Uh it's not really showing up in the numbers that way.

5:43

This is a very very different story.

5:46

Uh but I do think that the >> I'm sure we've ruined that guy's night >> right at right at the right at the start of the >> Yeah, [laughter] I think I I I think he took it with a grain of salt.

5:57

Uh >> we were also asking also asking for his opinion on different tech leaders. Um >> loves them all.

6:05

>> Loves them all >> equally.

6:07

>> Equally, >> which is negative 100% love. >> Yes.

6:09

[laughter] But uh the other interesting uh discussion downstream of this is just um >> uh lots of people are rallying around Jensen, not just in the venture capital and startup ecosystem, which was the first community to really rally behind Jensen.

6:28

Uh the employees are now rallying behind Jensen.

6:30

Of course, there are 42,000 of them.

6:32

So hard to say exactly uh what the what the approval rate is, but uh Zoe Schiffer over at Wired and Pesh Dave uh have reported that some Nvidia employees are calling for Jensen Wong's return even after his multiple reported extrammarital affairs and abrupt resignation.

6:50

According to internal correspondence reviewed by Wired, an internal suggestion box post submitted before Wong's resignation by a longtime engineer at NVIDIA urged the company's board to take employee sentiment into consideration prior to taking any action.

7:04

The submission has attracted more than 1,600 comments since then with several asking board members to find a way to bring him back.

7:13

If you believe in the mission, bring back the boss, all caps, uh, wrote one anonymous commenter.

7:19

So 1,600 out of 42,000 employees.

7:22

Uh this interestingly it's not just private comments.

7:27

This was a public Instagram post from someone who works at Nvidia. Bring him back.

7:32

Uh very clearly stating her opinion that he should be brought back.

7:37

I thought this was interesting because this account is normally focused on uh bringing engaging with the startup ecosystem and uh the person that runs this uh was using this opportunity to to to rally the troops.

7:48

Nvidia is nothing without its people.

7:51

A lot of a lot of OpenAI folks were saying go read the original letter that Meera and Ilia and everyone signed during this during the blip during the OpenAI blip.

7:59

And so uh these things can kind of flip back and forth.

8:03

There was some sk there was some speculation that he might be back at some point.

8:06

I was trying to get people to to uh weigh on um both seem unlikely, but Jensen at SpaceX running Terra Fab versus Jensen back at Nvidia.

8:16

Which is more likely between those?

8:19

I think most people would say the most likely scenario is, you know, sail off into the sunset, maybe do some investing, maybe start some philanthropy, maybe even start up, start a new company, robotics space, something interesting where you can be surrounded by engineers, but you're sort of out of the limelight.

8:34

Uh, if he were to go back to Nvidia or go to SpaceX, he would be very much back in the limelight.

8:40

Uh, those both seem low probability events, but um, which one do you think is more likely?

8:46

I think going back to Nvidia is probably more likely than working than being a founder for 30 years and then going and working as an employee effectively at SpaceX. I don't know.

8:55

At the same time, you know, Michael Traller wound up working at SpaceX.

8:59

>> Such a tough information environment right now.

9:02

>> It's all speculation at this point.

9:02

It's just people people guessing >> and you can see why employees want it back. >> Look at the stock.

9:11

>> Why shareholders want it back?

9:12

>> Yeah, look at the stock price.

9:12

the discount the the the you know the discount that the business is going to have without Jensen I think is undoubtedly going to be extreme. >> Yeah.

9:22

>> Um not all the impacts will be felt immediately. >> Yeah.

9:27

you know, for obvious reasons.

9:27

It's a it's a you know, unique culture and and you know, extreme talent density, but >> um >> yeah, it's this interesting thing where I feel like people are sort of normalized to extreme variance among individuals, right?

9:40

Like >> I'm sure, you know, looking at at Elon, you have this like some things that are just absolutely incredible about Elon and then some things that are maybe >> less ideal for >> demon mode. demon mode, right?

9:55

As Walter Isacson talks about or just like the >> in the bad column you put demon mode.

10:00

In the good column you put you put Diablo abilities. >> Yes. Yes.

10:05

Um >> but I think people are so used to understanding at at this point >> Yeah.

10:11

>> understanding individuals as like extremes, right? >> Yeah.

10:14

Mercurial >> and uh >> and all about tradeoffs. Yeah.

10:17

Well, let's go to Klet Crest's former employer, Cisco.

10:22

critical infrastructure from the AI era.

10:23

Unlock seamless real-time experiences and new [applause] value with Cisco.

10:27

Uh Jensen is on the cover of the Wall Street Journal again.

10:33

Uh taking shots at Colette Crest, the new interim Nvidia CEO.

10:35

Uh the headline before his ouster Jensen belittled the executive replacing him.

10:42

Interim CEO Klet Crest has questioned some Wong dealm.

10:46

He called her a quote deer in the headlights.

10:49

He said that uh she was more of a reporting CFO, quote, not someone who could help him define a strategy or make decisions.

10:58

And there's two reads on this. One is uh she was wrong.

11:00

She wasn't risk on enough during she was holding him back during the the greatest uh economic boom in this industry in history.

11:11

Um the other is that uh leverage is risky and that uh for such a profitable company uh it is more financially responsible.

11:19

It is doing a duty to the shareholders to be a little more risk off and not act as the central bank and not use as much leverage and not uh do do so many circular deals.

11:27

So there are two sides of this coin and we'll see uh what happens to Nvidia if they continue to run the Jensen playbook with a new CEO who's just as risk on in the dealm world and just as much of a fast fast action mover and shaker or if they go to someone like you know a Tim Cook or a or a uh a Colette Crest who would probably be a little more buttoned up a little bit more restrained in the deal making. Um, >> who fits that?

11:55

Who fits that deal making CEO archetype >> to come in? >> No, no, no.

12:01

I don't think there's anyone outside of Sam. >> Sam? Yeah. Who else is doing stuff? I don't know. It's hard. Yeah, it's rare.

12:10

It's very rare to be to be moving that quickly.

12:12

Um, I don't know who else would would be coming in.

12:14

I mean, Satia did to his credit like make that very weird deal with OpenAI early on and everyone was like, "What is this?"

12:21

And that panda under a blue well. Yeah, exactly.

12:22

Uh, so so it's not of pure, it's not only the domain of founder mode CEOs.

12:29

I mean, I would put Sate in the founder mode CEO even though he's not a founder.

12:33

Uh, he's not the founder of that company.

12:35

Whereas Tim Cook has been very restrained throughout his career at Apple, but the stock didn't suffer because of it.

12:40

And so there's a world where Nvidia switches into this manager mode, more committee-ledd, but they still manage to produce incredible growth, incredible cash flows, and the stock appreciates a ton.

12:52

And we're looking back at, oh yeah, the actual the the succession era.

12:54

Sure, it wasn't the the Steve Jobs era anymore.

12:58

It was more like the Tim Cook era, but Nvidia has performed incredibly well because they were set up for success.

13:04

Uh, and the manager that was able to come in delivered.

13:07

So, that's another possibility, but we'll we'll see. We'll see.

13:11

Um, the other interesting thing is there's there's been some debate over uh Jensen's character or or who he really is because he's doing a ton of press, a ton of events.

13:24

He's always around, always traveling.

13:26

Uh, he's also incredibly aggressive on deal making and doing all these deals, but then he's also a technical leader. He's also an engineer.

13:34

And so, uh, how much time can he really be spending looking at architecture of new uh, new chip designs?

13:38

We talked to Dylan Patel about that a little bit yesterday and I I I think I now understand it.

13:44

There's this piece of lore about Jensen that he has like 60 direct reports and no one-on-one meetings and and he's reading a thousand emails a day and it seems almost too much to believe based on the sk the travel schedule, all the other things going on in his life, especially with the new revelations.

14:05

But the interesting thing is that I do think that part of the founder mode purpose is to purely be there to break committees and that actually is not a very timeintensive piece of the job.

14:20

So, I think there's this world where an Elon or a Jensen can come in and and see and read the room and understand that lieutenants have been biting it out and working really hard and bringing their best to bear, >> making their case and you understand the history and the individual track record and and where their where the what their biases [clears throat] are and what their individual aims. >> Yeah.

14:44

And so even if you're even if you're sometimes not a 100% on the technical architecture because you haven't done all the research yourself and you're making sort of a gut call, you're going to be right because of the, you know, 30 years of experience.

14:58

You're going to be right like 80% of the time, but basically a 100% of the time you're speeding up the organization because you're like, "Look, we're going to disagree and commit.

15:07

We're going with Steve's plan or we're going with Colette's plan or whatever on this.

15:10

I I I want to do it this way and so everyone fall in line.

15:15

We're going full tilt on this thing.

15:17

And that's the that's the benefit of this founder mode aggressive CEO.

15:20

And that doesn't mean 80 hours a week working on that one decision.

15:23

It can be a 15-minute meeting or an hourlong meeting, but just having that authority in the room to say the buck stops with me.

15:32

I'm going to be dragged over the coals if this doesn't go right.

15:35

So you take the pressure off, but we're going with this plan.

15:37

We're going to move forward accordingly.

15:39

uh and so yeah h finding someone to fill that role uh is is very very difficult.

15:44

Um but yes, people are are uh are debating this uh left and right both in the in the founder CEO community and the venture capital community also with the employee base within Nvidia obviously.

15:55

Uh before we move on, let me tell you about Crowd Strike. Your business is AI.

15:59

Their business is securing it.

16:01

Crowd Strike secures AI and stops breaches more important than ever.

16:06

We have some more news on cyber security this today. So Toby look chimed in.

16:12

said generational blunder at least based on the information that's available and he's quoting the Nvidia newsroom which was announcing the CEO transition.

16:20

Uh Ronald Vice says 33 years as CEO and no permanent successor ready.

16:26

The blender started way before >> also this is just like an AI response.

16:31

>> Okay, >> this is like how they this is this is basically how AI bots respond on.

16:34

They just sort of like add some >> fact. >> Yeah.

16:41

than a more relevant real time fact.

16:44

And then like they their take, >> but it's not really a take.

16:50

>> You think this account is AI?

16:52

>> Yes, >> it has a blue check mark and 25 followers. Jordy, come on.

16:53

[snorts] That stretches the imagination. Uh, no. But Toby engages.

16:58

He says there's no replacing.

16:59

I don't care if you're AI.

17:02

>> Ratios him out out of the solar system.

17:04

[laughter] >> 13,000 likes.

17:05

Uh, take a >> silence bot. >> Silence bot.

17:09

Uh there's no replacing a founder like Jensen.

17:12

There has never been and there can never be a true succession for a founder like him. Nvidia ends today. What does that mean?

17:20

Uh and is replaced with a thing cosplaying Nvidia.

17:22

It will be run by custodians trying and failing to preserve the flame he ignited.

17:26

Uh and uh yeah, people are going back and forth in the comments to this.

17:33

The good news is the AI trade has a new potential bullcase >> which is >> Jensen coming back. >> Oh, [laughter] right.

17:43

>> I mean, >> you think >> it's like back to the back to, you know, the war.

17:47

It's like, >> oh yeah, >> the war started, but >> but that creates a bullish catalyst because if we end it, >> there there hopefully is bullish catalyst. >> Sure. Sure. Sure. Yeah, that makes sense. Um, yeah.

17:59

DHH it takes 10 years for the culture of a great company to fall apart once the CEO seat is given to someone without an engineering or product background.

18:06

That's been the story of Boeing, Intel, and now Apple.

18:07

Uh the degree to which it's over for Nvidia cannot be stated.

18:11

You cannot have been counterated uh cannot be overstated.

18:15

Uh you cannot have Bean counters in charge of a technical organization like that. Okay. Um yeah.

18:20

Um the interesting >> Yeah.

18:22

Here's the so here's here's my view on this is like um it takes 10 years for the culture of a great company to fall apart once the CEO seat is given to someone else. >> Mhm.

18:32

>> Uh to someone without an engineering or product background.

18:34

So the crazy thing about Jensen is he has that engineering background and the product background and also the like the deal making machine >> which is a huge factor in Nvidia having so much market share. Yeah.

18:48

the the uh your your belief that they're going to continue to maintain a huge amount of market share um that they're continue to make the the the best products in the category.

18:59

And so this is even like the potential impacts are not, you know, I look at like Apple where Apple now struggles to make, you know, new products that people really love that seem to work and and they feel, you know, kind of stuck in time, right, in the in the sort of like >> it's lighter, faster, thinner. >> Yeah.

19:23

>> More breakable than ever before.

19:23

Um and and it's sort of like playing out playing out Steve's legacy >> and it's the biggest company in the world.

19:33

It's a fantastic business.

19:35

>> Um but Steve wasn't playing the role of like >> the central bank of like an entire industry.

19:44

He wasn't doing this deal making.

19:45

making. And so the difference is that I do believe that is just like so important to that's like a real part of Nvidia's moat is like Jensen's dealm >> and so now you lose the engineering and product leadership that founderled

19:58

company and you lose the dealmaking and that's how you end up in just like a >> and and that's why every single you know shareholder that I'm seeing every sing you know all these employees are saying like bring this guy back he has his flaws. >> Yeah. He's shown poor judgment. But if >> Yeah.

20:13

He's shown poor judgment.

20:13

But if these investigations happen and there's no smoking gun and he was just selling computers >> Yeah.

20:23

>> to the best of his ability within, >> you know, the letter of the law, >> the letter of the law, then like I can see at this point >> in the same way that like the intensity of the story earlier this week by by Tuesday, it was like I can't really imagine >> him staying in the seat.

20:39

Now it's like, you know, start you're starting to see a world where where he could go back. >> Interesting. Um, yeah.

20:46

Another thing on the deal making side is that it feels like his his one of his advantages there is being the engineer and being able to clock technical teams.

20:54

like a lot of the investments that he's made.

20:57

Uh he's able to move very quickly on the finance side, but he's able to uh you know evaluate maybe this company is overvalued on a revenue basis, but uh are they building something real? Is the team there real?

21:12

And uh being able to being able to uh correctly clock uh the technical abilities of teams is really important in those like fast fast deal making scenarios.

21:22

Uh, Taylor Swift [clears throat] might be in the running.

21:26

Uh, Papas is reporting that Taylor Swift is making a big announcement today on Good Morning America.

21:30

And Paul Graham says, "Now I'm really starting to wonder about Nvidia's board's judgment." I don't know. Put her in.

21:36

I think she might be good. Who knows?

21:38

Um, we should move on because otherwise we'll just be here.

21:42

And we have Tam joining at 11:30 to discuss this more.

21:44

Uh, is there anything else you want to talk about with uh Nvidia before we move on?

21:48

Oh, last post >> from Hunter Spx Thompson.

21:52

He says, "All right, this is tough.

21:56

>> Huang had once found a soup vendor in Japan," he said with a long line of customers.

22:00

He joined the line and after a long wait, enjoyed one of the best bowls of soup he'd ever had.

22:03

When he asked for another serving, the soup seller told him no.

22:08

He had too many customers waiting and not enough soup for all of them.

22:12

"He's a better businessman than all of you," Hang told the cloud computing CEOs.

22:16

The soup man doesn't sell out all his inventory to one or two customers, but instead keeps prices high and his customer base diverse.

22:24

The CEO said, >> "Wow, >> you're not establishing the value of my products," he said.

22:29

>> Wow, that's [laughter] a great metaphor.

22:31

>> So, so elite because that's exactly what he's doing. >> Yeah.

22:34

>> Um and you know, encouraging his customers to do. >> I love it.

22:38

Uh let me tell you about Codex.

22:39

Codex is a powerful workspace for getting work done with AI [applause] agents.

22:42

Whether you're writing code, analyzing data, creating content, or automating business workflows, Codex helps you move projects forward with from start to finish. >> Triple glaze.

22:51

>> Triple [laughter] glaze on Codeex.

22:51

TSMC was hacked through an outside cloud provider that stored records on its current and former employees.

22:58

The company confirmed the leak after dark forms and breach forms listings offered 6. 9 gigs.

23:04

The seller says holds around 46,000 employee records.

23:07

The price is three Bitcoin. It's awfully cheap. Uh around $220,000.

23:15

Uh with a threat to dump everything free after a week unsold.

23:18

We should we should buy it and just mail them nice care packages.

23:23

Thank you for your service.

23:25

You know, if we're going to get all the TSMC employees, we could just send them nice Christmas presents.

23:28

You know, >> that's right. Or at least a card.

23:31

>> Put them all on the nice list.

23:31

Make sure that they're on the nice list.

23:34

>> So, do we know what is actually in the employee records? Is it compensation?

23:40

And is it >> 19 fields per person including national ID numbers.

23:43

So basically social security numbers, addresses, bank details, and emergency contacts.

23:47

So very very bad from a uh if this falls into someone who wants to attack TSMC employees.

23:52

Very very bad if you just want to uh create a bunch of fraud alerts and identity theft and all sorts of stuff.

24:00

I mean the corporate espionage angle. >> Yeah.

24:05

Is potentially really >> Yeah.

24:07

Then also, I mean, a lot of companies are starting to keep their broader employee bases a little bit more optimized >> because you don't ask for worked >> if you're going to ask for money. >> Yeah.

24:20

>> Like, who who's silly? Yeah. Yeah.

24:22

>> Who's silly enough to like I have all of your employees data and I will leak it to the world creating massive downstream implications for their safety and the, you know, safety of your corporate secrets.

24:33

And you know, I am asking for $220,000 [clears throat] because if you're TSMC, you're just thinking like, do we pay the ransom or not, right?

24:43

It's a very binary thing.

24:45

Like, if the ransom was like $100 million, I don't >> I mean, the trick is that like, you know, you pay the ransom.

24:50

Who knows if there's another ransom? It's always hard.

24:53

The game theory there is always tricky.

24:54

Uh, do we have an update to the naughty list? The nice list. We've been tracking.

24:58

>> I think we got to put these hackers on the naughty list. >> VPN list. Oh, we got a graphic.

25:00

to the naughty or nice list. Who?

25:03

Oh, Microsoft Outlook has been added to the naughty list. It went down.

25:08

It has been down for a while.

25:11

Uh, the latest direct from Microsoft. TLDDR. This is very, very bad. Uh, status. cloud. microsoft.

25:19

If you're checking up, there's service degradation on Microsoft 365. >> Is it still down?

25:25

>> And Microsoft consumer products.

25:25

Um, so some people are saying that they're able to log in, but not all emails are being transferred.

25:31

Tyler Cowan had a post on his blog, Marginal Revolution, that said, uh, "My GMU Outlook email is down.

25:40

Go figure out my Gmail if you need to get in touch with [laughter] me."

25:42

And he was like, "Or try Twitter." Uh, >> yeah.

25:45

Microsoft's uh status page is still saying service degradation. >> Yep. So, very very uh rough.

25:50

What uh Tyler, can you figure out the like the most significant outage of Microsoft 365 or Outlook in in history this has going going >> going >> we're now past 24 hours?

26:07

>> If you're Microsoft Outlook, Jordy, and you get added to the naughty list, what is your reaction emotionally if you're Microsoft Outlook, you get added to the nice list, to the naughty list?

26:18

>> Potentially a bullish catalyst, which is getting the service.

26:20

I think if you're Microsoft Outlook and somebody says you're you're down in the worst outage ever, there's only one thing to say.

26:27

>> There's only one way to go up.

26:28

>> There's only one word to describe what you're feeling. >> What's that? >> Good. >> Good.

26:33

[laughter] >> You have the opportunity.

26:36

>> You're really setting me up for that. >> The nice list.

26:37

If you're on the naughty list, you need to lock in. You need to say good.

26:43

>> You have the opportunity to make the nice list.

26:45

And if Microsoft Outlook comes back and there's no information stolen and they learn a lot, I'm putting them on the nice list. You heard it here first. >> Good. >> Yeah.

26:53

So, I believe this is the worst Microsoft 365 outage like ever.

26:58

>> So, I in 2023 there was one where it was down for like around 5 hours for in the worst case.

27:04

>> Is there any is there any news on like what's going on?

27:07

Are they under attack or is this just uh somebody was vibe coding and and shipped something accidentally to prod that they shouldn't have?

27:15

Like >> is it >> I don't think it's a cyber security breach or anything like an attack. >> Interesting.

27:20

>> Um yeah, but I don't know exactly the reason. >> Yeah.

27:23

Burning burning the the scale too much.

27:25

Too many maybe I I mean GitHub was going down because of all the agentic coding traffic.

27:30

Like it is possible that there's just higher load from agentic emails being sent across Outlook.

27:34

You have to imagine that a lot of Outlook addresses are the targets of of agentic effectively spam.

27:41

So, uh yeah, it might just be like a load balancing issue or something. I don't know.

27:45

Uh Will Depw has been digging into the top four Chinese AI labs, DeepSeek, Moonshot, Miniaax, Ant Group.

27:53

Uh he claims that they built their newest models on Nvidia's most advanced Blackwell chips, which are barred from import.

28:00

And so they shouldn't have been you.

28:03

these black wells should not be making it to China and yet he was digging into the to the training methodology the code bases.

28:13

Now Dylan Patel chimed in and said uh this is widely known because you can uh just look at the code and see that uh DeepS open sources it and if you just look at it uh for 10 seconds you can figure out that it's trained on Blackwell.

28:28

Um but Will uh is replying and saying like yes like but we need to be more explicit about this.

28:34

We need more uh investigation and analysis and this moves Dylan Patel's uh you know insider knowledge a little bit forward and a little bit more out into the open.

28:46

Uh what's very interesting is um just how small uh how how few black wells you need to train one of these models.

28:52

need to train one of these models. uh and and for a few reasons like the big labs have much more compute many more black wells but I think that when it comes to the the methodology of the American labs first they have to serve their models whereas the open- source uh Chinese models often get served from other clouds so you don't have to hold

29:13

back if you only have so many GPUs you don't need to be like oh well we got to save some for inference or there's a lot of inference demand for our previous thing we can't take that offline and then also uh a huge amount of training compute goes towards sort of exper uh experimental training runs where you don't know if it's going to work or not. If you're doing a distillation model,

29:29

If you're doing a distillation model, something that's uh where you're a little bit more sure that it's going to work, you actually wind up needing less compute to basically catch up or stay near Frontier.

29:40

So, uh interesting that you can that you can do this with uh you know le I don't think they're talking about hundreds of thousands of black wells going there.

29:49

It's probably a much smaller amount probably in the billions of dollars which of course uh makes the Nvidia story with uh with uh Mega Speed and the other affiliates uh even more problematic because it's it's it's a small amount of revenue relative to Nvidia, but it's been a huge thorn in the side of everyone who's been uh worried about the impact of of open source.

30:10

Of course, there are a lot of people that are excited about open source and so it's been a it's been a hot debate for a long time.

30:15

Well, if you're sick of our coverage of Go for it, John >> agents, meet the canvas.

30:21

Your AI [applause] agents can now create and modify your Figma files with design system context.

30:29

>> If you're sick of our coverage of the NVIDIA story, you're going to hate.

30:33

>> You're going to get sicker.

30:33

Get ready to puke cuz we got the Nvidia expert himself.

30:37

Take him in the TBPN Ultradome. We're bringing him in.

30:40

Welcome to the show, Tag. How you doing?

30:46

Thank you so much for joining the show.

30:47

>> Slow news week this week.

30:49

>> Slow news week indeed. I'm exhausted. How are you doing?

30:53

>> Uh it it's been rough. >> It's been rough. >> Getting sleep. >> Yeah.

30:57

>> That that night when uh the board was meeting, I had to like figure out do I get any sleep or should I stay up all night?

31:03

I knew >> a decision was going to happen and it was >> Yeah.

31:09

>> Yeah. Sometimes when you have a week of of uh like an insane news cycle, this has been a different experience just because it's like it's uh such a you know you can't look away but it's also incredibly tragic right for >> the technology industry for Nvidia for

31:27

Jensen and and >> um you know I think people are like actually seeming to do a good job of holding two things which is like people can be flawed and make mistakes and simultaneously be you know, one of one and incredible in in other domains. And

31:40

And so despite all the chaos of like, you know, insane over on Instagram is like crazier.

31:48

I'm seeing posts with like, you know, half a million likes that are just like I'm like, h this, you know, the story is not niche at all, right?

31:55

It's like everyone knows >> you know Jensen and all the they can recognize all the the other people involved but um >> overall I think people have generally been pretty balanced about it >> which I I guess is a silver lining. >> Yeah.

32:14

>> I I've been talking to people at Nvidia and it's complete shock. >> Yeah. >> Disappointment. They're angry. >> Yeah.

32:20

So many of the causes that Jensen um was a proponent of now those things are are in big trouble. >> Stranded. >> Yeah.

32:30

>> The the China hawk stuff, the open source movement. Yeah.

32:34

>> Like all these things that are legitimate uh positive >> Yeah.

32:39

>> forces like that's all at risk now.

32:39

And I just want to say like this is just such a sad ending for uh probably one of the greatest CEOs of American history.

32:49

Like to have this tenure end like this, all the stuff that he contributed to society, the technology industry and pushing AI forward, it's really it's really disappointing and sad. >> Yeah.

33:00

I have a hedge fund manager friend who's uh his his one of his big overarching thesis is that we are coming to the end of the boomer bubble.

33:08

Uh, and this obviously relates to the the handover of generational wealth from the boomers and boomer politicians going, you know, we've had boomer politicians for something like 40 years or something like that.

33:21

And uh, he was he was I he was calling out that uh, Jensen was the the final Boomer CEO to go and now we're firmly in like the Gen X millennial territory across uh, all the big tech companies essentially.

33:35

Uh, it's yeah, it's a it's a very remarkable time.

33:37

Uh, as you think back on, >> which is crazy because Jensen at no point in time have I mentally thought in my head, okay, Boomer, you know, >> he's like so >> so so much vitality, so so much so much energy.

33:54

Um I guess like what are you you know I would say by Tuesday I couldn't imagine Jensen staying despite that just with all everything sort of piling on despite that Nvidia would never be the same and there was going to be a bad bunch of bad downstream implications.

34:13

It just felt like like just too much in in some sense.

34:18

Now it's like I still think you have to believe there's a chance he comes back, but there's a lot of investigations. >> I I think it's over.

34:27

It's like I I tweeted it out as soon as this the second affair report came out with five sources.

34:34

Like >> yeah, >> I I knew then it was completely over because >> um >> this is direct report >> direct report and and a direct report is in the NVIDIA terms and conditions basically their employee handbook of conduct.

34:46

code of conduct >> and it's it's much more cut and dry, which is weird to say because we're talking about allegations of spying with a geopolitical rival, like the most controversial thing.

34:56

But there is that that investigation has not concluded. This one is cut and dry.

35:01

And so the second one was truly the end.

35:05

>> And the board must have known that if there's two Yeah.

35:09

>> two ants in the kitchen, as people say, there's probably more.

35:11

And then Reuters came out with their third one. >> Third. Yeah. >> Um Yeah.

35:14

I mean, it was just game over because the code of conduct says uh conflict of interest. >> Yep.

35:21

>> And ethics decision tests.

35:21

Like if this was in social media or on top of the New York Times, would you want to do this?

35:28

And obviously it did not meet those tests.

35:30

And um yeah, you can't you can't have an affair with a customer. Yeah.

35:35

>> That's a conflict interest.

35:35

And you can't have affair with your subordinates because you have power over these people.

35:39

And that's that's just game over. >> Yeah. Yeah.

35:42

No, and it makes sense if you're working at a company and you're the one not having the affair.

35:46

You're like, "This is not a level playing field for a company."

35:49

There's so many reasons why that that that >> when I was writing my book, uh >> yeah, >> I did over 100 interviews.

35:57

>> Uh it was focused on how video was built, how how how effective it was at business culture.

36:03

But I did ask, are there any negative behaviors I should worried about?

36:07

>> And no one said infidelity. Mhm.

36:10

>> Um, all the negative behavior, dozens of stories of him, all the bad behavior was him like berating executives and employees in meetings.

36:18

You're completely incompetent.

36:20

We're going to go out of business because you're so incompetent.

36:24

And that, >> you know, that's his frank, that's just his management style.

36:26

I saw the Wall Street Journal report and Colette, that's just what he does in every meeting.

36:32

It's not [laughter] it's not different.

36:34

That's his standard operating procedure uh for 30 years.

36:36

Um, so I kind of laugh that people are like going off on >> It's also standard operating procedure in a lot of companies, not just >> Oh, really? Okay.

36:46

>> Uh, I mean, yeah, the the like the going back to the the financial crisis, you hear about bankers and hedge fund guys like screaming at each other across.

36:54

>> John always likes to say the good part of getting yelled at on a conference call is you can turn the volume down. [laughter] >> Yeah.

37:02

>> So, the one thing I want to say is I did circle back I did circle back to some of my sources. >> Yeah.

37:07

And they I asked about Kimberly Powell >> and they said that um >> they didn't suspect any infidelity, but >> multiple sources told me they were kind of surprised how quickly she moved up the ladder, how quickly she got promoted. >> Interesting.

37:20

>> It didn't go through the normal process of that's that's really rough. >> Stuff like that.

37:25

You can't it's over like you can't deal with things like that. >> Yeah.

37:29

Can you give me any more color on there's now a debate.

37:33

I was talking to somebody who's who's very skeptical of tech, Jensen, Nvidia, all this stuff.

37:40

And uh and he was saying like, well, now that we know all this, like it's very clear that Jensen was just playing a character.

37:46

There's no way he was reading a thousand emails a day.

37:47

There's no way he was involved on the technical side. like it.

37:52

He was just the dealmaker and chief going around papering money all over doing podcasts and events and uh and he had a whole team that handled everything else.

38:03

How do you think about his role prior?

38:06

>> That's that's completely false.

38:06

>> That's that's completely false. I have dozens of firsthand accounts of him >> like one executive told me like I think this is my book he got like 15 20 emails a day and he knows all his other [snorts] >> executives his colleagues are getting that kind those kind of emails I mean

38:22

they're not like three paragraph emails >> it's almost like texts yeah just just what's the status on this what's going on this like give me an update why you doing this wrong it's those are the kind of emails uh he spends all weekend writing these emails you know with his laptop with his whiskey Sunday night. So, I mean it he's definitely

38:39

So, I mean it he's definitely doing everything.

38:42

He's the chief marketer and chief Yeah.

38:45

>> chief business strategist.

38:45

Like he micromanages people.

38:47

He's always looking ahead.

38:49

Um and he's HR and chief by recruiting, you know, the best talent from across the industry and >> retention.

38:56

So the question is if he is all these things and he's so great at all these things and he's proven he he was he's been able to do this in three decades, how is Nvidia going to survive? Yeah. >> With the >> Yeah.

39:08

Uh how how have you been thinking about succession planning internal, external?

39:13

There's been a bunch of names thrown around.

39:15

Some of them more technical, some of them more just tech legends.

39:19

I think didn't Donald Trump uh uh identify Tim Cook, which is sort of >> I just don't Cook.

39:25

I don't think [laughter] he's definitely not NBC.

39:28

I don't think Colette is she's a >> Have you seen the latest Tim Cook news?

39:34

Have you seen the latest either of you? Uh Donald Trump.

39:37

>> I I I don't think this was AI.

39:37

I might be getting fooled, but I saw uh a picture and some reporting of him in uh in like a European country buying a a very expensive rug from like a rug dealer.

39:49

You got to check this out, Tyler, and figure out if it's real or fake.

39:52

But it it was that the message was very clear.

39:54

It was like Tim Cook is enjoying retirement.

39:56

He's traveling the world.

39:58

He's building a rug collection.

40:00

He has no desire to get back in the most controversial battle in the world and play all that out.

40:06

And I and I tend to believe that even if the image was fake.

40:09

>> I I don't think the board would be interested in Tim Cook.

40:11

Anyway, so the two the two candidates I did put out there publicly are Renee O, the CEO of ARM.

40:17

He actually used to work at Nvidia, so the executive team knows him well and I know, you know, from sources that they loved working with him. He was always prepared.

40:26

Um so I think that would be a good fit and and he can play the public facing role uh business management and and let the internal Nvidia executives on the technical side really thrive autonomously. >> Sorry continue.

40:42

>> Oh and the other person I think is a good candidate is Ian Buck the creator of CUDA.

40:46

He's in charge of Nvidia's data center business has a customer relationships and he has Jensen's ability to see the technology future.

40:53

I I've been quoting that um there was this podcast in 2019 where he was talking about large language models, scaling laws, natural language processing, all the stuff that happened three years later.

41:06

>> Is that Ian Buck you mentioned? >> Yes, Ian Buck. Yeah.

41:09

>> So passing the buck potentially.

41:12

>> The buck stops with Buck. This works.

41:14

The nominative determinism I'm behind.

41:14

I don't know anything about this.

41:18

>> Buck money, you know, you can keep going.

41:21

>> I mean, I've interviewed him.

41:21

He speaks at conferences.

41:23

I think you'd be a great choice.

41:25

And oh, one thing I do want to mention is there are these guys uh become CEO, you have to lock down Colette and this guy named Jonah Alvin who does he's the head of EPU engineering >> throw a billion dollars at these people >> and they got to stay.

41:42

>> You have to lock down the executive talent because I'm sure you know a lot there's outside recruiters at other companies.

41:47

If you lock down the top, you know, two, three people and other Nvidia sources tell me actually there's like 10 people you need to lock down. >> Yeah.

41:55

>> If the board is able to lock down these people, I think Nvidia can be successful.

42:00

>> So, what what did you read into this Jensen uh this in the Wall Street Journal before ouster Jensen belittled his replacement interim CEO Klet Crest had questioned some of his dealmaking.

42:09

He called her like >> just said that's what he says to everyone. >> Yes.

42:14

But strategically, every meeting strategically, why is that?

42:16

Like, is this an effort to that Jensen doesn't believe Colette should be in the CEO role, but he's fine with her staying as CFO or there's actually something deeper there?

42:29

>> I I don't think well, I don't want to put words in Jensen. Sure.

42:32

Like, I don't think Colette is the permanent person.

42:39

That's the new people and I think they're taking their time a I mean there's definitely a short list and they're taking their time and Financial Times reported uh like an hour ago that they they are talking to Renee Hos. >> Okay. Wow. >> Yeah. Yeah. It's exciting. Sorry.

42:52

>> So, I think they they need to vet their backgrounds now because after what happened with Jensen, you really have to make sure there's no skeletons in anyone closet. >> 100%. 100%.

42:59

So my big question for you is how much of Nvidia's moat on a go forward basis would have been Jensen's dealm because there's a DHH has a post where he says it takes 10 years for the culture of a great company to fall apart >> once the CEO seat is given to someone without an engineering or product background.

43:21

That's been their story of Boeing, Intel, and now Apple.

43:22

Um, Nvidia is going to continue to make fantastic computers.

43:29

The question is like in order to maintain market share and and uh it it feels like Jen it doesn't it feels like you could find somebody with amazing engineering and product experience that's respected by engineers that has zero ability to do the the the real like high-flying deal side and that gets pushed to committee and Nvidia ends up suffering because of it and it creates room for for new entrance.

43:58

So, so, so, so what I foresee in the best case scenario is say Renee Hos uh becomes a CEO, he can do the conferences, he can do the marketing, he can do the business management and then autonomous let like Jonah Alban be autonomous on the GPU technology engineering.

44:17

Let Ian Buck manage the data center relationships and have his technology strategic vision.

44:22

um instead of one Jensen, maybe you have, you know, two two three people that can do parts of Jensen.

44:32

>> And like I said, the the most important thing that the the culture is ingrained there, like a mission is a boss, speed of light, like insane work.

44:39

All these people work 100 hour weeks.

44:41

The important thing is you need to hire a good CEO that knows his limits, his or her limits, >> and then make sure you retain these this deep bench that has been executing and knows the technology.

44:58

>> You need to retain these people because, you know, I'm sure every other company is calling these people right now. >> Sure.

45:03

>> If they're able to retain the top five, 10 people and have a good CEO, the two candidates I mentioned, I think Nvidia can do well.

45:09

I mean the next two three years are actually baked in the cake in terms of their act the quality of the products.

45:14

The key question is what's the strategy going to be the deal making uh three to five years from now and and that's the critical question the next CEO obviously the quality of CEOs is so important in the technology industry. >> Yeah.

45:27

>> What uh what do you think Jensen does next?

45:30

Still feels like uh it's hard to imagine him going out like this and then just being done like he >> I mean I think Well, I mean, first of all, he needs to go through this investigation process and go before Congress and, you know, >> he's going to be very busy on the legal political front the next few months or maybe even a year.

45:52

>> Um, I don't think he's retiring because like like he's famously said, work is the way he has fun.

45:59

So I don't I don't think he's done like >> as soon as all this political stuff and all this legal stuff that he has to go through and he's done say you know 6 to 12 months from now he'll do something.

46:12

>> Um I do think he still loves Nvidia like Nvidia is his life's work so I don't know if he'll compete with Nvidia but but we'll see. >> Yeah.

46:21

Uh well we found the post uh the picture of Tim Cook retirement side quest update. Can we pull this up?

46:27

Uh, Tim Cook spotted buying rug for $18,000 at the Grand Bazaar in Turkey. Do we have the image?

46:36

I think we put it in the timeline in the >> So, this is real.

46:40

This is >> It looks like a real image, but it's from uh EOG, Entrepreneurs on IG, which is I don't know if it's Wall Street Journal level coverage.

46:50

I'm not leaning on it, >> but uh it it is uh it is a good indication of where Tim it's believable even if it's fake news. We'll see.

47:00

>> I mean, after all the work and all the stuff he he's he's gone through, >> like, leave the man alone.

47:05

He's he did a great job for Apple.

47:07

[laughter] He doesn't need to be the center of geopolitics and running another major technology company.

47:15

>> How are you expecting earnings to go next week?

47:20

Um, so I do think, okay, there's a feeding frenzy now. Yeah.

47:26

>> Uh, because Jensen is 247 news and the politicians are just, you know, they love that stuff, so they're going to get ahead of it.

47:32

So, I do think on the margin, >> uh, >> the data center loophole where, you know, Chinese customers can access Nvidia products on on these international data centers, I think that's gone.

47:44

They're definitely going to pass legislation uh to ban that.

47:45

pass legislation uh to ban that. um that's been the loophole that people go I think so on the margin I think uh parts of the business it's it's not huge uh are going to go away so I don't know what Colette's going to do um I do think there's no impact in terms of overall

48:05

aggregate AI compute demand it's overwhelming every single data point shows there's massive demand >> for uh GPU compute >> but on the margin I do think that that loophole is going away and and and on the margin maybe Neato clouds might have a tougher time uh financing their capex over the next 12 months. >> Yeah. So I mean we've been hearing >> Yeah.

48:25

So I mean we've been hearing rumors about uh private credit sort of slowing down a little bit in some of the longer tale of the buildout.

48:33

Uh but you know you had those great coreweave earnings.

48:37

There's a whole bunch of different data points around uh some positive stories.

48:41

there's some uh mag seven earnings misses and big sell-offs and then and then building back from there.

48:48

there. But it feels like uh I mean all eyes are on the on the earnings uh things this should not be this should not be uh derailing from business as usual but even if none of this was happening there's still a question of you know expectations are so high have their have the jitters in various pieces of the market slowed down demand at all so I think people probably scalers I think are going to be able to raise

49:13

whatever they need to raise to hyperscalers >> have cash flow on the margin I think the financing environment for neo clouds has become more difficult and there's a lot of macro reasons for that with the fed actually hiking rates and

49:27

>> the financial conditions tightening um and a lot of fears about the election that it's going to be harder for these companies to get data center permits and etc but I don't think it's as of yet that significant Yeah. >> And it's more of like an incremental on

49:41

>> And it's more of like an incremental on the margin type of thing versus like being like 30% of revenue is going to get crunched because of all these different factors.

49:49

And I do hope um that the war ends soon because I do think that's a major impact on inflation.

49:54

Like if if we can get diesel prices and oil prices cut in half because the war somehow >> uh comes to resolution and I think >> politically Trump can't escalate after what happened in the midterms. >> Yeah. No way.

50:09

You would think that in within weeks or months hopefully the war winds down, they come to something.

50:14

Y then energy prices, oil prices will go down.

50:18

The Fed won't have to, you know, this morning one of the Fed uh presidents like threatens like being much more hawkish >> on CNBC. Um that goes away.

50:25

The Fed doesn't have to be super hawkish and then that that's going to let finance companies like Neoclads raise uh money more easily.

50:36

So there's a lot of macro crossurrens, but I don't think it's like dramatic and significant yet.

50:43

>> Yeah, Jordan, I have one more. >> Go for it.

50:46

>> Um, how how are you thinking about net blackwells or net GPUs flowing to China?

50:56

This feels like it will decrease compute supply in China in the short term.

51:00

This feels like an event, a catalyst that will widen the gap between closed source and open source. Is that your framework? Is that your your prior?

51:11

>> Well, I first I mean there's two things I want to say.

51:13

First, China everyone like all this discussion, China is the one that's blocking Nvidia GPUs from getting there.

51:21

Like H200s, the the US government finally got to a point where okay, we'll give licenses.

51:25

And the reason why Nvidia can't sell, you know, GPUs to China is because China was offended by Howard Lutnik's, you know, we won't have to go into it, but sure, he was very insult insulting them and, you know, China decided to block H1 200.

51:41

So, >> yeah, >> it's China that's blocking Nvidia GPUs right now.

51:44

>> So, but but is that true?

51:44

Because I mean we saw this Will Depw tweet go out and he's he's been digging into all the different uh >> No, you need to be able to understand there's two things happening that there's the posture of the government and there's what the labs want to do.

51:58

The labs just want >> the best compute they can get their hands on >> however they can.

52:04

>> You're talking about the labs being Deep Seek, Moonshot, Minia Max, and >> Yeah.

52:07

They're just like we want compute. >> Yeah. Yeah.

52:10

>> But the government is taking a different policy.

52:12

And so you can't look at it as a single stance. >> Yeah.

52:17

So I would assume that if they close that data center loophole, they stop shipments to Malaysia, they stop shipments to Singapore, they stop diversion, they stop direct shipments to China.

52:27

If those are happening, the the the total amount of compute available, Frontier compute available to DeepSec, Moonshot, MiniAX, Ant Group, decreases over the next year or two.

52:37

Now you could have uh Huawei and SMIC and SME like take off.

52:43

That's certainly a possibility but at least in the short term it feels like uh China is going to be more GPU poor in the next year or two.

52:51

>> So I think the bigger problem for for China's models is not that it's that right now we're going through this RSI thing. Yeah.

52:59

>> Where open AI are hitting a part of the >> uh R&D cycle where >> AI is just going to take off.

53:06

And what both these labs have done over the last I think this has been undercovered >> is that they've been cutting off the distillation accounts. >> Yeah. >> Yeah.

53:19

>> And I think what you're going to see in the next six months is that the models from anthropic and open AI are the the gap between anthropic open AI and versus Chinese models are going to widen dramatically. >> Yeah.

53:33

>> Because the only reason I mean this is obvious if you've been following everything.

53:37

The only reason these Chinese models are good is because they're just distilling the American models. Yeah. It's not like that.

53:42

That's >> And they're buying and they're buying the same data. >> Yeah. A little bit of Yeah. Yeah. Yeah.

53:51

>> It's mainly constellation and if you if Anthropics has been very good at this kind of, you know, putting out a document that that shows what's going on. Yeah.

53:58

And both these labs have been clamping down on it.

54:01

And as RSI takes takes effect, I think the the leadership of the merit and the AI labs are gonna really show up. >> So yeah. Yeah.

54:10

So all of that seems like uh even with all the turmoil and the chaos in with Nvidia, but also the market broadly, uh it's not the craziest time for Anthropic to go public next week.

54:23

It feels like they reset the valuation and it's perfect, right? >> Yeah.

54:27

and the October AR which is I mean I've said this other places I think part of open strategy of like really putting putting pressure on anthropic from >> July to September. >> Yep.

54:40

>> Uh kind of backfired because it made anthropics numbers look worse in September >> and then >> Exactly.

54:47

And then Opus 55 came out.

54:47

is such a great model that their growth >> the way their growth curve looks from the September quarter to October looks great and Wall Street is not that smart like they're kind [laughter] of like goldfish they just you know >> what happened this month. Yeah. >> Yeah.

55:04

So if if you see what happened in the last week, all this sentiment like everyone's freaking out, anthropic IPO is not going to happen and then someone >> very brilliantly at the bottom of the market leaked the October ARR saying it was going to be over 80 billion to to Wall Street Journal and then sentiment completely flipped around. Yeah.

55:25

>> And now we're all there, seven times. >> Yep.

55:28

Six six, seven times >> uh overs subscribed.

55:30

Um they set the thing low and now it's going to go great.

55:34

and the the the IPO is going to go bonkers next week.

55:36

So So that's going to help sentiment. >> Yeah.

55:39

Yeah, that makes a lot of sense.

55:40

>> And this RSI thing, by the way, is very bullish for AI semiconductors because >> the more advances you get, the more chips you need to buy to to accomplish the RSI. >> Yes.

55:52

But it feels like it's bullish for uh semiconductors broadly.

55:54

The the semiconductor igopoly, it's good for TPU, tranium, Intel, AMD, uh the startups.

56:02

Yeah, >> agents and GPUs.

56:04

And you know, I've been interviewing uh the head uh chip designer at OpenAI. >> Yeah.

56:13

>> And all they say is we're taking everything we can get.

56:18

>> I talked to a meta a meta executive at conference.

56:23

>> We're just taking anything we can get, but there's overwhelming demand. Yep.

56:27

>> For they can bring them up faster with agents.

56:30

they can actually get things running on the on different pipelines.

56:33

>> The the open AI chip design head chip architect tell told me uh the agents has been a dramatic multiplier and accelerate. Yeah.

56:42

>> In terms of their chip design.

56:44

>> If you had to write another book, who would it be about? Would you do AMD Intel?

56:49

>> I I did I did tell people I'm working on a second book.

56:51

I'm not ready to >> You're not ready to disclose it? >> Disclose it. Yes.

56:55

>> I think we might need 10 more books. We need a TPU book.

56:57

We need a tranium book, an Intel book, an AMD book, a uh maybe a Broadcom book. >> I'm not an ARM book.

57:05

>> I I can only write one book at a time. >> Get working, Ty. Get working.

57:07

We need these books faster than you can write them because uh the the semiconductor race is fully fully on.

57:15

Uh we are we are potentially leaving the monopoly era and so many interesting >> interesting every week. >> Every week. Every week.

57:22

I mean, we didn't even talk about the midterms and the elections that that much and it's just incredible. >> Yeah.

57:29

>> Uh, every week is like a brand new crazy story. >> Yeah. Yeah. We were talking. >> This is exciting. >> Yeah. Yeah. Yeah. The Yeah.

57:34

It'll be interesting to see you the the the >> full employment for Tay Kim.

57:39

>> Full employment for anyone with a tape.

57:42

It's a It's a great It's a great time. And >> Great to see you.

57:44

Hope you can get some rest this weekend. >> Get some rest. Happy Friday.

57:48

>> Next week will be interesting.

57:49

>> We'll talk to you soon. >> Cheers. >> Goodbye.

57:52

Let me tell you about MongoDB.

57:52

What's the only thing faster than the AI market?

57:56

Your business on MongoDB. Don't just build AI.

58:00

[snorts] Own the data platform that powers it. We were close. We were close there.

58:06

>> Uh, our next guest is in the waiting room already.

58:08

We got Ben Smith from Semaphore, the co-founder and editor and chief.

58:13

Ben Smith, how are you doing? Happy Friday. Welcome to the show.

58:17

[music] >> Nice to see you.

58:18

I am I'm overwhelmed with news, but uh what a week.

58:23

>> Everyone's having uh indigestion.

58:25

>> You're mostly focused on GTA launch next week and kind of leaks around that or [laughter] or >> we're we're locked in on on that.

58:31

Yeah, thanks [clears throat] for asking.

58:34

>> Top there's nothing else going on.

58:36

>> Nothing else going on.

58:36

H how do you think I mean we've been talking about the Jensen news all all week.

58:40

How do you think this changes uh tax relationship with DC?

58:45

How do you think things will will shape up over the next couple weeks?

58:51

>> Yeah, I mean I think you know it's funny because we're we're hosting this conference ne next week which is now incredible time and it was sort of whose premise was well you know what like policy makers and technologists ought to be talking to each other more and now it just really feels like they're on a collision course >> and now it just feels like they they crashed the cars crashed.

59:07

Um, and you know, I I mean, I think in some ways one way to interpret it is that that the Silicon Valley figures who made a big bet on the Trump administration, >> Jensen being kind of like one of the central ones are now paying a price for that. >> Yeah.

59:24

you know, and because in a way I think if you put aside the alleged the kind of alleged spying and you put aside the sex, which I realize are a lot of things to put aside, what you have is this story about chip exports and about the Trump administration kind of deciding that >> it seems that they were going to let a certain amount of gray market chip exports go into China.

59:44

And it's an administration just doing that stuff unilaterally. >> Yeah.

59:48

um and and between Jensen and Trump and and and Howard Lutnik and a few others and and what you're seeing now feels like kind of the revenge of of the system of the national security bureaucracy.

59:58

So, I can't I can't fully remember, but I'm trying to trying to think if I It's so easy to visualize a bunch of different moments where Jensen was next to Trump or on stage somewhere just singing the admin's praises and, you know, more more so than pretty much any other tech CEO that I can think of.

1:00:22

Like even with with Zuck, who you know started to look like at least right-wing coded with some of his interests, right?

1:00:29

MMA and wakeboarding and all these things.

1:00:31

The only thing I can remember with Zuck is him being like kind of maybe overpromising on the capex side.

1:00:37

It felt like that at the time, but now he's actually delivering it.

1:00:41

>> Whereas with Jensen over the last year, it's like, yeah, really, really, really just >> heavy heavy heavy support.

1:00:45

And it felt like just not a politically smart thing to do, right?

1:00:51

Maybe he didn't need to go that far.

1:00:53

But in hindsight, the Mega Speed story first broke over a year ago.

1:00:58

Alice Hong was in the news a year ago.

1:01:01

It almost feels like in hindsight maybe that just like over o over o, you know, overly praising the admin.

1:01:05

It was like a reaction to the beginning of the story, which we still don't have a lot of clarity on, right?

1:01:11

and sort of speculation that uh you know Trump had his 60 Minutes interview.

1:01:15

There's speculation that you know the per the AI leader that that Trump was saying take off that and then he paused and he said and and get ready to put on an orange suit speculation that that was you know about Jensen.

1:01:30

And so maybe the overwhelming praise was was just part of this whole story and something that he felt like he needed to do to to keep the game going. >> Yeah.

1:01:42

Yeah, that's that's I kind of I mean that is a fascinating theory honestly because I think so much is being interpreted and I think the way I've mostly seen it is look how much Nvidia got out of the administration >> in terms of just opening these markets you know and and there was a real

1:01:57

alignment of interests like I don't think you necessarily need to imagine that there's more like Trump just loves the big investment numbers he sees this as you know he sees and particularly for first years saw you know the central element of his p of his presidency. Um

1:02:09

Um just just you know huge investments in AI and Nvidia you know in some you know right at the heart of it.

1:02:17

Um I mean but I think one thing you know that you see that Congress really really wants to know about is Jensen getting blocked from the plane Trump texting and then saying meet us in Anchorage and then jumping on the plane.

1:02:30

Like that really feels like what what whatever it is that happened that is a metaphor for it.

1:02:35

happened that is a metaphor for it. Um and and that that image of him on the tarmac in his leather jacket I think is going to be kind of the indelible image of this of what of this mess which is you know it's a huge Washington mess like it is you you have your sh there are always scandals there's news cycles and just it's been a while since one

1:02:54

just kind of escaped containment like this and is you know has sort of the there's you know going to be committees on the hill looking into this there's going to be probably internal investigations on on the administ ration side it's totally entangled with the financing and the infrastructure of the AI boom like I don't really see this this news cycle slowing or stopping. I

1:03:14

I think we just we actually just published an amazing story in SE4 as I speak about some of the uh some of the inner workings but it is like it I don't remember a story this big.

1:03:23

It's been a while like it it has slightly blacked out the sun of the midterm elections which were a pretty big story. >> Yeah, >> totally.

1:03:30

I >> no I was >> for me Jensen's been CEO of Nvidia my entire life >> and it's never felt felt like there was this monumental of a of a story throughout that entire period. >> Yeah.

1:03:44

>> At least that that I can remember.

1:03:47

>> I'm still going back and forth.

1:03:48

>> Is there Nvidia without Jensen?

1:03:48

What's your >> What is your view on Nvidia?

1:03:50

I think it's immediately a shell of its a shell of its former self and uh >> you know it's tough enough to pull a founder out of an organization.

1:04:00

The the culture is very strong.

1:04:03

the culture will endure for some amount of time.

1:04:05

endure for some amount of time. But uh everyone that talks about Nvidia on a go forward basis is saying well well this person will be this part of Jensen and this person will play this other role that Jensen played and >> you end up all roads to me lead to like you know running a company by committee and what has made Nvidia so successful is that like you know effectively

1:04:34

running it like dictatorship style willing to make >> you of fast decisions willing to do the sort of like I just I don't believe there is a CEO out there that can do the deal making and and navigate the supply chain relationships and all of the

1:04:49

downstream customers, the hyperscalers, the neoclouds, the application companies, open- source and then also do all the engineering and and product leadership that that Jensen was so good at. So like um completely irreplaceable

1:04:59

So like um completely irreplaceable and that's why I think you see calls from shareholders and employees right now to bring him back even though there was like a very severe you know code of conduct violation that that is inexcusable in pretty much every >> I think I I I firmly disagree with this.

1:05:20

I think if you bring if you find the Tim Cook of Nvidia to Jensen Steve Jobs, you can see serious appreciation in the value of the business. >> I agree.

1:05:30

>> I agree. I didn't say I didn't say there's no world like like Apple's >> like market cap has increased dramatically and I think that that's possible but I still would that the the the soul of the company >> agree agree and it's and it's naturally >> whatever you know think about I I still believe if Apple if if if Steve Jobs was running Apple and you still had Tim Cook

1:05:54

doing his thing yeah >> you might be looking at a10 trillion dollar company right now and I think the same thing for Nvidia it's like Yeah, >> the stock price can appreciate, the company can get bigger, they can continue to be a beneficiary, but >> it just, you know, 5 years out from now, 10 years out from now, there will be a a Jensen discount. >> And and to be fair, I mean, it seems

1:06:10

>> And and to be fair, I mean, it seems like when Steve Jobs passed the torch to Tim Cook prior to his uh his passing, um he had gifted Apple a stronger monopoly than Jensen gifted Nvidia.

1:06:23

I would say that the Apple monopoly is stronger than the and the Apple monopoly >> because it's like a market share the market share is flipped. >> Yeah.

1:06:32

>> Like Steve's gift wasn't the like I'm giving you, you know, the entire market. Yeah.

1:06:38

>> But it was I'm giving you a real monopoly. Yeah.

1:06:39

That's going to come >> on the app store and that will just produce forever. >> Yeah.

1:06:45

>> And and the Apple analogy is interesting, right?

1:06:46

Because also and the critique of Apple since Jobs is they stopped innovating, but also were super focused on profits and margins and a lot of things Wall Street likes.

1:06:54

read a story today on that just, you know, the trade-off.

1:06:58

I mean, innovation's expensive and the Wall Street doesn't always like it and Nvidia was spending a lot of money on on big bets and things like that that a more disciplined company may be a company that at least for a while the Wall Street likes better even if it innovates less. >> Yeah.

1:07:10

Um, [clears throat] how do you think about the scale of what Nvidia got from the administration with sales to China?

1:07:20

because it's always seemed small relative to Nvidia.

1:07:24

I mean, they're big numbers.

1:07:26

It's like $8 billion here.

1:07:26

Mega Speed was doing billions of dollars, but Nvidia has $60 billion of cash on its balance sheet, a hundred billion sometimes.

1:07:36

I think the the the the buyback program is $250 billion.

1:07:38

And over the last two years, like during the Trump administration, even if you include all of Singapore, all of Hong Kong, all of Malaysia, you're you're only looking at 40 billion of revenue.

1:07:52

It's not that much at the Nvidia scale, which is crazy to say because that's a ton of money, but uh it is it is marginal in the Nvidia ecosystem for a $5 trillion company. >> Yeah.

1:08:03

And and what's more, of course, there's there going to be they're willing buyers for those things.

1:08:05

There's a lot of people awaiting the big news obviously of the season with the GTA release who um who who would like a nicer GPU and can't afford it. >> Yeah, that's true.

1:08:15

Yeah, >> it's not like it's not like right.

1:08:16

It's not like if the sort of Singapore based smuggling companies went out of business, nobody would be buying that stuff at all. There's huge demand.

1:08:24

>> I mean, you hear that story Yeah.

1:08:27

of like of like Elon and Larry Ellison begging Jensen for GPUs.

1:08:28

I don't know how real that exact story is, but like there's clearly insane demand labs and and the hyperscalers like you could sell those chips in America. >> Yeah. >> Yeah.

1:08:41

And they may be winners in the short term, right?

1:08:42

Like it's be easier for OpenAI to get chips. >> Totally. Totally.

1:08:46

>> I mean, I think but you know, I think the Jensen would had this sort of bet on world domination and on particularly dominating these Asian markets on I guess also in the medium-term trying to run ahead of of a huge Chinese investment competitors. >> Sure. Sure. Yeah.

1:08:59

And I mean, if you play it out, if he's thinking in 30-year time horizons and you look at the uh the energy production and the energy curve, the energy generation curve in China, uh they will have vastly more power than America.

1:09:14

Uh and if if if that becomes the most important bottleneck in AI and super intelligence, well then if you're if you don't have a business there, uh you are truly not in the game.

1:09:26

And so yeah that what that that that global thinking that that that world domination mindset uh leads you to a strange place for sure. For sure.

1:09:36

>> How how have you process like the way in which the story broke uh quite a bit you know bombshell but quite a bit different than let's say you know other major stories you know looking at therronos as an example you know one one uh sort of key figure head. >> Yeah.

1:09:55

>> Uh journalist John Keroo. >> Yeah. on.

1:09:57

>> Same thing with Ronan Pharaoh with the Harvey Weinstein story.

1:09:58

This story doesn't seem to have a hero journalist, at least that I'm aware of.

1:10:05

>> Um, you know, I would I don't know if he's a hero, but Dimmitri Seastulo from the FT is somebody you always ought to read. Okay.

1:10:10

And and there's a bunch of people at the New York Times who broke pieces, but it has broken in a much more Washington way, right?

1:10:14

Like the most classic Washington way, which is Marco Rubio wrote a memo. Yeah.

1:10:19

>> And like whenever some whenever in Washington you write a memo and then you give it to like 10 people and they all know that that is a handwritten, right?

1:10:25

Like you get that memo and you're like, "Oh crap, now I've seen it." Okay.

1:10:29

>> And at some point somebody's going to leak it because a memo like that is written solely to leak essentially. Got it.

1:10:33

>> And to sort of cover cover your ass.

1:10:33

And so um at some point, you know, Rubio months ago distributes this memo.

1:10:38

And it's that sort of sets a ticking time like sets a clock on when is this thing going to leak.

1:10:45

And then the way it leaked was totally in the terms essentially that Rubio had set like this woman appears to be a spy.

1:10:50

There's an inappropriate relationship. National security risk.

1:10:53

security risk. it now seems like the appears to be a spy thing is a little more complicated and there's some descent inside the intelligence community of exactly what her relationship with Chinese government is and also she's not like a >> you know movie Red Sparrow agent type

1:11:08

she seems to play like kind of a different role in the Chinese business ecosystem so I don't know like it's a but it definitely like came out of Washington at 1,000 miles an hour >> with a narrative that was like really clearly set by by you know one side of the argument and I think everyone is sort untangling that. Now, it's kind of

1:11:23

Now, it's kind of striking how little the administration has even really even tried to get out its its point of view here.

1:11:29

Like obviously at some level like Trump and Lutnik investment like they all knew that >> well David Sax did sort of cut out because they could have read it in the New York Times. >> Yeah.

1:11:41

I mean, David Sax did post sort of a line by line rebuttal of the spy allegations on Alice Wong, saying that uh various pieces of the investigation are either inconclusive or dispositive uh towards exactly what role she was playing.

1:11:57

And he kind of outlined a little bit more of what you described there about maybe more of an asset than a spy, someone who wasn't necessarily trained internally. >> Don't worry about it. Just an asset. >> I don't know. Yeah.

1:12:06

Well, it's a difference between so but but so it's sort of primarily your Is your primary job you check in at the spy facility, the spy job, or or or are you a real business person who just takes phone calls from the the spying agency on occasion?

1:12:21

These are two separate designations.

1:12:23

Confidential source or whatever is what they call it in America, I think. >> Right.

1:12:28

>> Right. So that so that and but I think the other I mean just sort of the big I think if you put that aside it's just like all you know the administration knew that these companies were smuggling chips to China and they knew it because

1:12:38

they could have read it in the New York Times or on X or a million other places like we all kind of knew it and there has just been this question of okay like so why are you letting them do it >> and it's still remains a bit of an open question. I mean, I think you saw

1:12:48

I mean, I think you saw somebody was quoted today saying, "Well, it's um it's, you know, that the um that it's it's in the context of the ne bigger trade negotiations with the Chinese and the last time they put things on the entities list, it blew up the whole relationship and so maybe they're letting a certain amount get through as in some tacit part of that bilateral relationship with China, but that hasn't really been explored."

1:13:11

Um, I mean, we have a big story that I'm hoping published in the last 5 or 10 minutes about the tight tight relationship with Nvidia and the bureau that was supposed to be overseeing it, the extent to which Trump was just kind of steamrolling and and Wong and Trump were just steamrolling the bureaucracy.

1:13:28

But I think there's just so much more still to come out. >> Yeah.

1:13:30

Uh, what else do you want to highlight at uh, Semaphor's new Silicon Valley and the World Summit? What else?

1:13:37

Is there anything outside of AI that people want to talk about?

1:13:39

Do they want to talk about bio yet or things that are at least like adjacent to AI?

1:13:43

Uh and then also I'm interested in what how else do you expand the world because again the only two countries you've mentioned so far on this 15inut show is uh is China and the United States.

1:13:56

Is there anyone else that's interesting or relevant these days?

1:13:59

relevant these days? Yeah, I mean I think the the idea of this this event that we're having Monday and t next week Silicon Valley in the world is you know is that policy makers everywhere are obsessed with Silicon Valley and you know Silicon Valley leaders often don't

1:14:15

love to go to Davos and are pretty busy >> and like but and so they we sort of said to a bunch of the leaders that you know at where you are like if we bring these policy makers to you >> you know is is their interest in having these conversations and so yeah I mean

1:14:28

we'll have the president of South Africa seros So we've got um the Mistral CEO from from France um is the leading European AI figure along with the you know some European leaders the uh >> the um the prime minister of Estonia will be there and so so I think we're >> they they love tech and then Pacilica

1:14:48

which is sort of the US-led sort of attempt to build a kind of alliance we'll have ministers from the UK from Taiwan from all over um you know and I think obviously though the questions are about AI they're about um and they're about [clears throat] Anthropics IPO, right? Like what a week to do this

1:15:03

Like what a week to do this honestly.

1:15:05

Like I I feel very lucky that it's all just like converging. >> Yeah. Yeah. Lots to talk about.

1:15:08

Well, >> and like you said earlier too, GTA >> GTA 6 most >> GTA 6.

1:15:17

Uh well, thank you to remote in and do a panel on GTA.

1:15:21

Maybe we can figure it out.

1:15:23

We would have loved to have you, but you >> have a special we have a special guest that day. >> Yeah.

1:15:28

But and also it's Thursday, so uh we we won't have too many spoilers [laughter] for for GTA 6. Yes, >> Jensen. We're doing a live play. No questions.

1:15:39

Just try and beat the main story immediately.

1:15:44

>> Have a great one, Ben.

1:15:44

Thanks so much for coming on.

1:15:46

>> Good luck with all the prep and I'm sure it'll be a great Ben. >> Goodbye.

1:15:49

Let me tell you about public.

1:15:51

Invest across all asset classes.

1:15:54

Automate anything with AI agents.

1:15:54

run training strategies, risk management, money movements, and monitor markets with AI agents for investing on public.

1:16:04

Up next, we have someone from the Wall Street Journal.

1:16:06

Gunjan Banner is the markets reporter and host of the WSJ Money interview.

1:16:11

[music] I'm sure you've seen these all over social media.

1:16:13

We have some crossover in our guests.

1:16:15

Tons of crossover in our interests.

1:16:18

Thank you so much for taking the time to join the show. How are you doing? >> Good. Thank you for having me. I'm excited to join you.

1:16:26

>> Thank you for joining.

1:16:26

Uh since it's the first time on the show, give us a little bit of your backstory.

1:16:29

How long have you been with the Journal?

1:16:31

What were you doing before?

1:16:32

What led you to uh the host, the Wall Street Journal money interview?

1:16:38

>> Yeah, I've been at the Wall Street Journal for nine years now, which seems crazy to >> overnight success.

1:16:44

>> We love an overnight success. >> Congratulations. >> Exactly. Thank you. It's funny.

1:16:47

I started as our options market reporter before options trading became cool during the meme stock mania and I've gotten to really broaden out, write about so many different markets topics, >> um, including the recent AI madness. Yeah.

1:17:04

>> And just this year I started hosting the WSJ Money interview which has been such a blast.

1:17:08

Got to travel around the country to sit down with um the coolest and most interesting tycoons of many different industries. >> Yeah.

1:17:18

what what are you noticing in terms of trends uh that are intergenerational?

1:17:21

Like we've been talking about the Jensen story.

1:17:22

He's sort of the last boomer CEO in tech, but I'm sure you're interviewing lots of folks who are Gen X or millennials.

1:17:29

Like what have you learned about either the way different eras, different generations of of successful business people think about money, think about wealth, think about luxury?

1:17:39

How has that uh changed or evolved over your reporting?

1:17:45

There's one big thing that really [clears throat] sticks out to me and that's applicable especially now where >> I think the most successful investors, entrepreneurs, they don't shy away from periods of market madness and I think a lot of them have actually made their money during crisis.

1:18:01

Um, you know, a lot of them were doubling down during the financial crisis, during the COVID pandemic. >> They were not afraid.

1:18:09

In fact, they were putting money to work.

1:18:11

They were buying build buildings. they were buying stocks.

1:18:14

So, I think that's especially applicable right now when you have all this AI craziness to remember that I think the people who've been really successful in markets, they really pounce on periods of crisis and turn them into opportunities.

1:18:28

>> Yeah, we yeah, we saw that a little bit with uh situational awareness, extremely dramatic story.

1:18:32

Uh Ken Griffin comes in and then a few weeks later we're hearing that Leopold's back in the market and and uh and running the strategy again.

1:18:41

Um what uh >> also Ken had been had some choice words that were >> what do you say >> leaking out prior to >> Oh yeah >> capitulation >> expecting a rate hike is that >> yeah yeah sure sure >> so understands the market cycle. >> Yeah.

1:18:57

Do do you uh speaking of Ken Griffin uh I he I believe he's a dinosaur collector.

1:19:01

Do you have you been seeing an upswing in dinosaur collection through uh the folks that you talk to?

1:19:08

We're we're huge fans of it here.

1:19:10

Haven't purchased a dinosaur skeleton myself, but maybe in the future.

1:19:13

Something to strive for certainly.

1:19:16

>> So, I haven't spoken with too many dinosaur collectors. Okay.

1:19:18

I will say I think collectibles are becoming huge.

1:19:20

I mean, look at all of the money that people are spending on things like baseball cards, >> basketball cards.

1:19:26

Like recently, I heard about someone paying for a basketball card that was tens of thousands of dollars.

1:19:32

>> So, I think collectibles are really huge.

1:19:33

I always love asking people how they are spending their money.

1:19:36

And I got to say that one thing kind of came up in a lot of my different conversations, which was a lot of these ultra wealthy investors, entrepreneurs, they love a million dollar vacation.

1:19:48

So, a million dollars spent on a month in Europe or a few weeks in Europe.

1:19:55

Um, >> you won't be surprised to hear that private jets came up several times.

1:19:58

And I actually had a really fascinating private jet mogul, Ken Ricky, on our show.

1:20:05

>> And he talked a little bit about how Yeah, I think he's been on here too.

1:20:09

>> Yeah, he's been on here too.

1:20:10

[clears throat] And he gave you a fantastic scoop about his tipping strategy, which is crazy.

1:20:14

He shows Do you remember the details?

1:20:17

It's like something crazy. >> It's It's epic.

1:20:20

And to be honest, I've started using it.

1:20:23

He says he tips when he gets there.

1:20:25

I'm not tipping as much as him, unfortunately, but he says he tips when he gets to a restaurant.

1:20:29

So, the first time he >> gets to a restaurant, he'll tip say pay $1,000 to the manager.

1:20:34

Um, the first time he gets to a hotel or a vacation rental, he might tip $10,000 when he gets there.

1:20:41

And he says that way people know what type of service or hospitality you you expect.

1:20:47

And >> I I love this abundance mentality about that. >> Yeah.

1:20:52

>> Yeah. I I worked at a I worked at a hotel in college and uh I've always I through that I realized like when you're going to any like specifically valet or really any service valet in particular if you tip when you come in you're going to get an amazing

1:21:07

>> service you often they'll keep your car >> at the top close by and uh often times the real hack is most valet >> don't actually charge you for parking if you tip them on the way in cuz they're just going to keep your car >> and then it's just faster because you don't have to deal with the thing. They

1:21:23

They know who you are and stuff. That's great.

1:21:27

Um uh what is how how do you spend a million dollars uh on a vacation in Europe?

1:21:33

I imagine that [laughter] uh getting there on a Gulfream back and forth that's expensive, but that's not a million dollars necessarily.

1:21:39

Uh what what's the breakdown? What are people doing? What does it look like?

1:21:44

>> Oh, what a great question.

1:21:44

Um and and I asked I asked the same thing.

1:21:46

I think private jet travel is a big part of it, right?

1:21:51

But then I also heard that yachts for several weeks can can cost a million dollars.

1:21:55

Like think about how the prices of hotels have surged.

1:21:58

You know, the journal had some great reporting how >> the number of hotels charging $1,000 a night.

1:22:05

These are not even, you know, the Amans of the world.

1:22:07

These are sometimes three or four star hotels.

1:22:11

>> It has just surged over the past few years.

1:22:13

So, I think you can get vacation rentals that are charging 5, 10, 15, $20,000 a night.

1:22:18

Um, so that that's how it adds up.

1:22:21

And I think the people I spoke with really viewed these as family experiences, like right investing in time with their friends and family.

1:22:27

And that's part of the the broader luxury travel boom we've seen the past few years.

1:22:33

>> Yeah, it it does seem like there's an entirely new class of hotel that's up at like the $5,000 a night range. That is something new.

1:22:39

uh all part of like this Kshape that's developing in so many categories.

1:22:43

Are there any areas that aren't K-shaped yet?

1:22:45

Is there an opportunity to K-shape something?

1:22:48

I I've always I've always laughed about the fact that like uh you know me me and Jeff Bezos probably watch the same movie.

1:22:55

It's not like he has access to like a hundred million dollar movie that no one else has seen or like we use the same devices. I don't know.

1:23:01

Maybe actually he does own he does own Amazon.

1:23:06

Think about like an album that one person has that the rest of the world doesn't and you can really only think of like >> Yeah. with the Wuang Clan.

1:23:14

>> He says, [laughter] >> he doesn't like it.

1:23:15

But uh is >> think about the the members clubs though, right?

1:23:19

Like you say is is anything not K-shaped and I feel like things are just getting more and more K-shaped.

1:23:25

I live in New York and just the number of members clubs that have popped up is is insane.

1:23:30

Um >> so yeah, it seems like things are heading that way. That's interesting.

1:23:34

Uh what what have you noticed any patterns with uh the the new AI boom specifically?

1:23:41

Uh changes in preferences around luxury or or or the way tech people spend money.

1:23:48

Obviously, there's a liquidity wave coming with labs going public and whatnot.

1:23:52

But I'm interested in anything that you've seen that uh we we we were talking about um the the collectible cars.

1:23:59

often it's like the poster that you had on your wall as a child and that used to be uh like a Lamborghini Countach, but for many millennials it was something much later uh and and something that was probably from 2005, maybe a Carrera GT or or a 918 Spider.

1:24:17

Um but what are you seeing among the the young AI wealth that's being developed?

1:24:22

Is that is that shaped differently?

1:24:26

>> So what I would say it's it's still really early there, right?

1:24:28

And I think we're going to see a ton of it with the anthropic IPO the next few weeks.

1:24:31

The Journal has had some really great reporting on how, you know, this young technology money has reshaped real estate markets around the country, right?

1:24:43

In in SF, I was talking to a friend in San Francisco who was telling me how impossible it is to find like a rental because people are just swooping in offering to pay hundreds if not thousands of dollars more a month for for a rental.

1:24:55

Um, I think it's showing up in travel.

1:24:58

Um, I was talking to a private jet executive who was like, I can't get my hands on enough to meet the demand.

1:25:05

Like there's not enough planes on the planet right now or under construction to meet the amount of demand that we have and and the wealth that we're going to have.

1:25:15

>> I also think some of these are um tried and true wealth indicators.

1:25:17

I was chatting with like a YouTube executive who had a big exit recently and he got a Rolex, right?

1:25:24

So I think some things kind of are popular among many different generations. >> Sure. Sure.

1:25:29

Uh when when you go into an interview, how do you think about uh uh the pre-work that goes into dialing in the level of questions so you don't necessarily cross a line.

1:25:40

I feel like when somebody sits down for the WSJ money interview, they're there to talk about money.

1:25:46

They're there to throw out some numbers, but there still might be you're not just going to say, "Hey, show me your bank account." Right?

1:25:51

Uh, how do you think about getting a good interview out of someone that is about a somewhat sensitive subject, but they know that they're going into a place where they're supposed to be a little bit more open than they might be in other scenarios.

1:26:06

>> I like to be really upfront about it.

1:26:07

You know, we have a no surprises rule here at the journal and I kind of try to adopt that going into the interview.

1:26:12

I mean, I would say it can be a little bit of a dance.

1:26:16

And you never want someone to be uncomfortable, but I I like to tell people kind of even when we're pre-ining or when I've first met them or when a a source first tells me, "Hey, this person would be really great for the interview," I like to say like, "Hey, we we also want to hear about how you approach this."

1:26:33

I think there are a million markets and money and finance podcasts.

1:26:39

I think not enough people are attaching real numbers as it pertains to wealth um into these conversations.

1:26:44

And so I I tell them I think there are really ambitious WSJ subscribers who want to hear about this.

1:26:53

And >> some of them are game and and some of them aren't.

1:26:56

And you know that's not always numbers based.

1:26:58

Uh Ken Ricky who who you guys had on like one of his viral moments was talking about these family money meetings that he has with these conversations.

1:27:07

And I can't tell you how many people reached out to me afterwards saying, "Hey, we learned so much from that."

1:27:12

Like, we had a family money meeting.

1:27:14

And I can't imagine a better response.

1:27:17

And that's what I'm hoping we get out of that cander.

1:27:19

And and I say that up front.

1:27:21

I'm sure you guys have have something similar, right?

1:27:25

When you guys are walking into interviews and telling people what to expect or where to go with something. >> Yeah.

1:27:30

Uh I I think um I I'm interested in those viral moments.

1:27:34

How how do you think about uh the fact that I feel like the WSJ Money interview particularly performs well in short form?

1:27:42

Like the clips do very well.

1:27:45

There are other Wall Street Journal shows that maybe don't maybe they're just more in the weeds, more detailed, and you can't get a one minute or two clip.

1:27:53

Do you think about that in the way you ask questions and the way you design the interview and the way you edit?

1:27:59

H how how integrated are you into the the the the editorial process that goes into the distribution of the content once it's actually been recorded?

1:28:09

>> I feel like we should have you in like our audience strategy meetings moving forward.

1:28:13

>> That'd [laughter] be great.

1:28:13

I'm happy to talk about this stuff all the time. I love I love content.

1:28:18

>> You have you have some really great observations about this.

1:28:22

>> I will say I don't think you can manufacture a viral moment. I wish you could.

1:28:26

I just I don't think you can.

1:28:26

And I there are moments where I'm sitting there in an interview, I'm like, "Oh my god, I cannot wait to share that with the world.

1:28:35

I think people are going to love it." >> Yeah.

1:28:38

>> And then there are moments we put things out on social and I'm like, "Really?

1:28:41

That got millions of views?" Like, >> I had no idea.

1:28:43

And you know, one of those moments for me was >> Venode Kosla um talking about how he doesn't get stressed.

1:28:50

[laughter] >> And I was like, "Okay, this is incredible.

1:28:53

Um I wish I did not get stressed.

1:28:56

I cannot wait to share this with my followers, the journal subscribers.

1:29:00

>> But it just went so viral on social and I could never ever have predicted that.

1:29:05

So I think the best that you can do is try to have a candid conversation and try to cover ground that has not been covered before >> and you know have have an illuminating conversation with nuggets that people have not heard before.

1:29:19

And that's the best formula I've come up with so far for success on wsa.

1:29:24

com and success [clears throat] on on social. >> Yeah.

1:29:29

So you don't want people sitting there saying I pay myself first, right? Isn't that the cliche?

1:29:32

[laughter] >> Yeah, I guess >> you want a little bit more detail, a little bit more meat on the bone.

1:29:36

Uh well, thank you so much for coming on the show.

1:29:40

>> Yeah, great to finally meet and congrats on on the run.

1:29:42

>> Yeah, it's been fantastic.

1:29:42

I I I look forward to the next episode dropping.

1:29:46

Have a great day and have a great weekend. >> Awesome. Same here. Thanks, guys. >> Talk to you soon. Goodbye.

1:29:49

Let me tell you about Console.

1:29:51

Console builds AI agents that automate 70% of IT, HR, and finance support, giving employees instant resolution for access requests and password resets. Good. Good.

1:29:59

And up next, we have Abashek Milani from Sequoia Capital.

1:30:04

He's a partner on the no limit part poker bot tournament.

1:30:07

We talked about this on the show.

1:30:13

>> Very excited to dig in deeper. How are you doing? Welcome to the show.

1:30:17

>> Thank you for having me.

1:30:17

First time on TV pan and excited guys.

1:30:20

>> Very excited to have you here.

1:30:21

>> Fantastic to have you.

1:30:22

>> Uh take us through the poker bot challenge. How is this going?

1:30:24

How did it come together?

1:30:26

How many people are competing?

1:30:27

Walk me through the high level. >> Definitely.

1:30:30

So call it about a month and a half ago.

1:30:34

Was playing a lot of poker with friends realizing these new models are coming out and >> holy crap they can do a lot of things.

1:30:41

And so we started testing over one of the weekends.

1:30:42

Can we build a poker bot?

1:30:45

So poker bots have been around for over almost a decade now, but you've never been able to do it quickly.

1:30:49

It used to be like a class for a full semester at MIT.

1:30:52

Then it was like weekl long competitions.

1:30:54

We started doing it in a weekend and I was like students way smarter than me can probably do it a lot faster.

1:30:59

And so come into partner meeting at Sequoa one day telling people this is something interesting that's now possible and immediately Sean Alfred a lot of the partners like we need to do something about this. >> Great.

1:31:11

And so when you think about poker bots, the first person you think about is Nome Brown.

1:31:14

He's the one who started this with his research at CMU in 2017.

1:31:20

So first call was Gnome and OpenAI. Yeah.

1:31:22

>> And they were excited.

1:31:22

They're like, let's see what our models can do.

1:31:26

>> And then we've been great partners with Citadel Securities for a while.

1:31:27

A lot of their job is thinking about strategies with hidden information.

1:31:31

And they've been incredible partners.

1:31:35

>> And from there, like month and a half ago, idea to finals is going to be this Sunday.

1:31:39

It has been an insane sprint, but over now I think 7,000 people have applied.

1:31:45

We have over 700 competing.

1:31:48

Finals is going to have your top 20 teams from both Boston, the Bay, Pittsburgh competing.

1:31:53

And we're pretty fired up of strategies. >> Yeah.

1:31:58

Um, >> what are like what are all the different approaches that people are taking?

1:32:01

I imagine some of the 700 is just like the most thin rapper of all time. just what should I do?

1:32:07

>> But then other people are like, you [laughter] know, >> asking the GPT4 like, hey, >> yeah, >> what should I do?

1:32:14

>> Give it your best shot, buddy.

1:32:15

>> Your [laughter] best shot. >> You're one.

1:32:17

Most people would quit right here.

1:32:19

They'd walk away from the table, but you know, 90% of gamblers quit right before [laughter] they're about the whole secret is a is a motivational prompt.

1:32:31

[laughter] >> Uh, this is also true.

1:32:31

I think the first versions of poker bots a couple years ago, you'd see these stories of the bot that shoves all in every hand just ends up winning. No one knows.

1:32:41

>> No one knows what to do.

1:32:42

>> Uh unfortunately, if you push in all in every time, you almost always come close to last.

1:32:48

Students have gotten a lot smarter, models have gotten better.

1:32:50

But >> that's what 99% of that's what 99% of gamblers say just before at the lower end.

1:32:58

You see some people trying to honestly like ask codeex or claude, hey, make me a really good bot.

1:33:02

I have to win or download all this hand history.

1:33:07

>> And then you get some really creative strategies and that's the bulk of the competitors honestly where >> they're thinking about the coolest one to talk through. >> Yeah. >> Yeah. The metag game. Exactly.

1:33:14

And each variant at each regional is different because we want it to be a new challenge every time.

1:33:20

Finals is going to be its own thing.

1:33:21

But at CMU, the team that won first place built this crazy classification model.

1:33:26

So, you come into a game and you start to figure out as fast as possible, is this player going to be shoving all the time, are they loose? Are they tight?

1:33:34

And within about like 15 hands of data, they start to figure out what is the Jordy bot going to play like.

1:33:42

And if you're going to be super loose, I'm going to go super aggressive into you when I have a good hand and try to exploit you. Got it.

1:33:47

And so, you start to see a lot of these strategies.

1:33:50

Some people make neural networks.

1:33:51

It's been pretty crazy the different types of strategies and exploits people have been making and it evolves over time.

1:33:57

The first version is different.

1:34:00

>> Are are there limitations on uh like on compute?

1:34:03

I forget the original rule set or could somebody come in and just apply like a massive amount of test compute to >> 10 billion [laughter] of compute to create the >> like is there a situation where somebody spends like a hundred grand on on >> No one No one would spend $10 million to win a $1 million prize in the modern era. Oh, that's that's crazy. No, no lab.

1:34:24

>> There's a lot of aura. You don't know. >> That's true. That's true.

1:34:25

The mill that was the story of the millennium prizes, right?

1:34:28

Like those were those were EV negative but great aura. [laughter] >> Exactly.

1:34:33

But no, there's a constraint where we have every bot has to respond within one millisecond.

1:34:37

This like prevents you doing like 30 minutes of test time compute before you make a decision and you have to be able to be >> control under constraints. >> One millisecond. That seems really fast. Wow. Okay. Cool. Yeah.

1:34:49

Yeah, but then I it allows you to actually play the whole simulation out very quickly. That makes sense. Got it.

1:34:55

Um, what about tool >> who So, so I'm uh I'm curious like one of the one of the interesting parts of this initiative for you guys and and Citadel and and OpenAI >> is that, you know, you're going to have, you know, potentially hundreds of like, you know, very very motivated, high agency, intelligent uh people.

1:35:16

How do you guys divide and conquer on bringing them into, you know, funding them, hiring [laughter] them, things like that?

1:35:26

Uh, or is it just going to be a free-for-all at the end?

1:35:31

>> I mean, I think at Sequoia, the core primitives of the business are people and markets.

1:35:36

This is a great way to bring incredible people together.

1:35:38

And >> maybe I'll give a >> some of the students like when we're filtering on like who are competing they're building Tesla coils that shoot 3 ft sparks.

1:35:48

Some are building their own PCBs.

1:35:50

Others are just developing their own like trading softwares that they've been doing in their high school PAS.

1:35:56

It's pretty crazy like what students are capable of.

1:35:59

>> And then in terms of like SEOA, we would love to be in business with founders from as early as possible.

1:36:04

But whether that is founding a company, going to do research at OpenAI or trade at Citadel or join one of other portfolio companies, it's honestly trying to figure out how can we best support this new generation and that's been the core mission from the beginning.

1:36:21

>> Longterm long-term view.

1:36:22

>> How do you think about how do you think about tool use generally?

1:36:24

I was watching uh some viral chess Instagram account uh do these very funny uh chess examples where he plays against the models or has the models play each other and they're just hallucinating crazy things.

1:36:38

So, it'll just be like an extra queen spawns over here and then and then it moves like a like a like a knight all of a sudden and there's just like insane things happening.

1:36:47

And I was like, this feels like a very odd model because in the modern era, if I go to any any frontier model with a proper harness, it'll just fire up Stockfish and just use a real chess engine.

1:36:59

It doesn't need to hallucinate the board or even think about it like it's solved.

1:37:03

And so, uh, in in in designing these, like how do you think about I mean, it it sounds like you answered it with the one millisecond response time, but, uh, are there open- source poker bots that the the models can build on top of? Is that relevant?

1:37:17

Is that discouraged or encouraged?

1:37:18

How do you think about that?

1:37:21

>> Yeah, I think the 1 millisecond is one thing, but the other core part is the variance themselves.

1:37:25

variance themselves. M >> so if you want to do standard no limit poker there's a lot of resources available online that will help you get you a good starting point but >> one I think crisp example in the Boston uh regional here's what the variant was you start with your standard two cards

1:37:43

but then instead of a flop of three cards you have a flop of five cards and you're actually bidding against all the other bots on the table for which three out of those five cards are going to be the actual flop >> and so this is a completely new mechanic that changes how you're going to play. If you just download a Stockfish version

1:37:58

If you just download a Stockfish version equivalent for poker, it just doesn't know how to respond here.

1:38:03

And choosing the flop is so important in the end outcome where you need to be able to develop the strategy in the tournament in order to really succeed.

1:38:13

>> Are you competing or are you recusing yourself?

1:38:16

Have you ever built a poker bot?

1:38:18

Give us your >> He said he did over the weekend that inspired it. >> Yeah. Yeah.

1:38:22

First time I built it, but unfortunately not allowed to compete given I know all the secrets. >> Okay. Yeah.

1:38:28

[laughter] Well, uh, congratulations and, uh, we're very excited.

1:38:32

>> Can't wait to can't wait to see the outcome.

1:38:34

>> We'd love to for the winner.

1:38:34

I mean, we'd love to have the winner on if, uh, if it's possible.

1:38:38

>> I I I want to play the winner bot. I I think I got a shot.

1:38:40

[laughter] >> I No one They've never seen anyone with so little experience in the training data.

1:38:46

[laughter] >> I'm terrible at poker. I never play.

1:38:50

>> Anyway, I'm sure it'll be a lot of fun.

1:38:52

Thank you so much for coming on the show.

1:38:53

Have a great weekend and we'll talk to you soon. Have a good one. >> Cheers. >> Goodbye.

1:38:58

>> Let me tell you about Shopify.

1:38:58

Shopify is the commerce platform that grows with your business and lets you sell in seconds online, in store, on mobile, on social, on marketplaces, and now with [clears throat] AI agents.

1:39:11

And let me also tell you about the New York Stock Exchange.

1:39:13

Want to change the world?

1:39:16

>> Raise capital at the New York Stock Exchange. >> There we go.

1:39:20

We got to get some more of these uh soundboard going. >> Yes.

1:39:25

Not everyone always says the show's I [laughter] enjoy the show, but not enough soundboard.

1:39:30

>> Not enough soundboard.

1:39:31

>> I think that's it for today, folks. >> Yes. Fantastic. >> Been a crazy week. >> Crazy week. >> Touch grass. >> We were sick. We got through it.

1:39:38

We're very excited that you all are here with us and I'll hopefully have some merch.

1:39:44

Uh thank you for tuning in to TBPN. We'll be back on Monday.

1:39:46

I think we have an early show Monday.

1:39:48

We're a Tomorrow X on Tuesday and Thursday.

1:39:50

Uh uh Tuesday and Wednesday.

1:39:53

Thursday is GTA and I think the anthropic IPO should be going out that day.

1:39:57

That's going to be a wild story. So uh tune in.

1:39:59

Don't forget to subscribe. Crazy timing. >> Ring the bell.

1:40:04

>> Anthropic IPO happening on International Men's Day. Yes.

1:40:07

It will be a very surreal moment for I think Daario and you know many of the other >> You think it's about International Men's Day?

1:40:13

I thought it was because it's Taigga's birthday on next Thursday.

1:40:14

I think they were kind of like, "Happy birthday, Taigga. >> Here's an IPO. Here's some liquidity.

1:40:22

Maybe you're an early investor." You don't know.

1:40:24

There's the hidden cap table.

1:40:25

There's the He could be in the line of successors, right?

1:40:26

Under the founders, there's a hidden list of people that succeed.

1:40:30

Tiger could be one of them.

1:40:32

>> You know, in a multi-layered SPV.

1:40:35

>> He also might be in a multi-layered SPV. So, you never know.

1:40:37

Could be a little >> You never know where where that guy's going to pop.

1:40:40

>> Tiger, you never know.

1:40:40

But we'll be covering it.

1:40:42

So, leave us five stars on Apple Podcast and Spotify.

1:40:44

Sign up for our newsletter at tbpn.

1:40:46

com and we will see you on Monday. Goodbye. >> Cheers. >> Goodbye. >> Have a great weekend.

1:40:53

Have >> the best weekend ever.